Proc. Natl. Acad. Sci. USA Vol. 93, pp. 2570-2575, March 1996 Biochemistry

Inhibition of the association of RNA II with the preinitiation complex by a viral transcriptional (/repression/immediate early 2/ ) GARY LEE*, JUN WUtt, PERCY Luu*, PETER GHAZALt, AND OSVALDO FLORES*§ *Tularik Inc., South San Francisco, CA 94080; and tDepartments of Immunology and Neuropharmacology, Division of , The Scripps Research Institute, La Jolla, CA 92037 Communicated by Robert Tjian, University of California, Berkeley, CA, October 17, 1995

ABSTRACT Transcriptional repression is an important start site of the HCMV MIEP (7-9). More recently, it was component of regulatory networks that govern expres- shown that purified IE2 protein binds directly to the crs sion. In this report, we have characterized the mechanisms by element (10, 11) and represses transcription of the HCMV which the immediate early protein 2 (IE2 or IE86), a master MIEP (12, 13). In this work, we have employed a transcriptional regulator of human cytomegalovirus, down- purified in vitro transcription system to study the precise regulates its own expression. In vitro transcription and DNA mechanism by which IE2 represses transcription. Our obser- binding experiments demonstrate that IE2 blocks specifically vations show that IE2 does not impede recognition the association of RNA polymerase II with the preinitiation by TFIID, yet very effectively blocks recruitment of RNA complex. Although, to our knowledge, this is the first report polymerase II. to describe a eukaryotic transcriptional repressor that selec- tively impedes RNA polymerase II recruitment, we present MATERIALS AND METHODS data that suggests that this type ofrepression might be widely used in the control of transcription by RNA polymerase II. Purification of RNA Polymerase II and General Factors. The RNA polymerase II fractions used in the in vitro tran- Substantial progress has been made in the identification and scriptions (DEAE 5PW step) and in the DNA binding exper- functional characterization of eukaryotic transcriptional acti- iments (Ila form, alkyl-Superose step) were purified as de- vators (for review, see ref. 1). In contrast, the understanding scribed (14). TFIIA and TFIID were purified as described (15). of transcriptional in has received con- TFIIF, TFIIH, recombinant TBP, IIB, IIEp34, and IIEp56 siderably less attention. Studies on transcriptional repression were purified as described (16-20). in have led to the hypothesis that eukaryotic Expression and Purification of Recombinant IE2 Proteins. repressors might function by a competitive mechanism, in Both the full-length IE2 cDNA and a DNA fragment encoding which the DNA-bound repressor blocks access of the basal IE2p40 were cloned into a pET-15b vector (Novagen). The transcription machinery to promoter DNA. Perhaps for this DNA fragment encoding IE2p4O was isolated with the use of reason, one of the first eukaryotic repressors to be recognized the polymerase chain reaction (PCR). The conditions used to was the simian 40 tumor antigen, which represses tran- express and purify the full-length IE2 protein and the IE2p40 scription by masking the promoter and inhibiting its recogni- protein were identical. Briefly, cultures (BL21) were grown to tion by the general transcription machinery (2, 3). The precise an OD600 of 0.5-1.0 unit and induced with 0.5 mM isopropyl step at which preinitiation complex formation is blocked by the j3-D-thiogalactoside (BRL) for 2 h at 37°C. Bacterial extracts simian virus 40 tumor antigen is unknown. The were prepared, and the proteins were purified according to the P-element transposase (4), the cellular DNA binding protein conditions described in the Novagen His-Bind buffer kit LBP-1 (5), and the bovine papilloma virus E2 protein (6) protocols. The affinity-purified proteins were then fraction- represent additional examples of repressors that block binding ated on a Mono Q column (HR5/5, Pharmacia). The molar ofthe preinitiation complex to DNA. In each case, it was found concentration was calculated by assuming that these proteins that the repressor blocked basal (TF) IID exist as dimers in solution. binding to the TATA box, suggesting that the first step in In Vitro Transcription Reactions. Transcription reaction preinitiation complex formation was the target for inhibition mixtures contained TFIID (DE-52, 1.0 ,ug) or recombinant by this class of eukaryotic repressors. human TATA binding protein (TBP) (60 ng), recombinant Important contributions to all aspects of eukaryotic gene human TFIIB (20 ng), recombinant human TFIIE (100 ng), regulation have come from studies aimed at understanding the TFIIF (TSK phenyl, 200 ng), TFIIH (400 ng), RNA polymer- coordinated regulation ofviral . Excellent viral ase II (DEAE 5PW, 100 ng), and TFIIA (DE-52, 5 jig). TFIIA model systems include the dual-function /repressor was not used in reactions reconstituted with TBP. Reaction proteins encoded by herpesviruses. Human cytomegalovirus conditions were as described (17). Reaction mixtures were (HCMV) is a 13-herpesvirus that has become a common and incubated at 30°C for 45 min or as indicated. very serious pathogen in immunocompromised patients. In Gel Mobility Shift Assay. The protein components indicated this report, we have studied the mechanism by which the in the figure legend were incubated with a 3'-end-labeled DNA HCMV immediate early protein 2 (IE2) represses transcrip- fragment (80 bp) containing the core promoter sequences of tion of its own promoter, the potent major immediate early the HCMV MIEP (bp -34 to +5) and polylinker sequences of -promoter (MIEP) of HCMV. Transient transfection the plasmid Bluescript SKI (Stratagene). Reaction conditions experiments indicate that repression by IE2 requires a cis- regulatory DNA element referred to as cis-repression signal Abbreviations: IE, immediate early protein; HCMV, human cytomeg- (crs) located between the "TATA" box and the transcription alovirus; MIEP, major immediate early enhancer-promoter; crs, cis- repression signal; TF, transcription factor; TBP, TATA binding pro- tein; TAF, TBP associated factor. The publication costs of this article were defrayed in part by page charge lPresent address: Signal Pharmaceuticals Inc., 555 Oberlin Drive, payment. This article must therefore be hereby marked "advertisement" in Suite 100, San Diego, CA 92121. accordance with 18 U.S.C. §1734 solely to indicate this fact. §To whom reprint requests should be addressed. 2570 Downloaded by guest on September 29, 2021 Biochemistry: Lee et al. Proc. Natl. Acad. Sci. USA 93 (1996) 2571 were as described (16) and TBP (3 ng), IIB (1 ng), IIF (20 ng), works effectively with TBP alone, we do not believe that IE2 RNA polymerase Ila (alkyl-Superose fraction, 10 ng), and functions by attenuating a TAF-mediated activation step. This IE2p4O (2 fmol) were added as indicated in Fig. 6. The reaction interpretation is consistent with the fact that the assay we have products were separated by electrophoresis on a 4% polyacryl- utilized represents a minimal activator-independent reaction. amide gel [30% (wt/vol) acrylamide/0.8% N,N'-methylene The experiments described above show that we can recapitu- bisacrylamide] containing 2.5% (vol/vol) glycerol and lx TBE. late transcriptional repression by IE2 with the use of a purified in vitro transcription system. Repression by IE2 Is Dependent on the Position of the crs RESULTS Element. Because the crs element is located between the IE2 Represses the HCMV MIEP in a Transcription System TATA box and the transcription start site, it is plausible that Reconstituted with Purified Factors. To characterize the the mechanism of repression by IE2 might involve steric mechanism of repression by IE2, we developed a purified in hindrance with the basal transcription complex. To investigate vitro transcription system that supported accurate transcription this possibility we asked whether the position of the crs initiation from the HCMV MIEP. Simultaneous addition of element with respect to the TATA box was important for IE2 resulted in repression. We analyzed the effect of IE2 on HCMV MIEP basal factors with recombinant full-length constructs in which the distance between the crs element and significant transcriptional repression of reactions driven by the the TATA box was increased by 5, 10, or 15 bp (Fig. 2). To wild-type HCMV MIEP (Fig. 1B, lanes 1-10). The level of analyze these promoter , we employed run-off assays. repression increased in proportion to the amount of IE2 As shown in Fig. 2, IE2 retained the ability to repress tran- protein added. In contrast, IE2 had little effect on transcrip- scription when the distance between the crs element and the tion reactions containing a HCMV MIEP with four TATA box of the HCMV MIEP was increased from 10 to 15 substitutions in the crs element (Fig. 1 A and B, nt (TATA 5 crs), indicating that IE2 does not require a specific lanes 11-15). These observations are consistent with previous stereo alignment with the TATA box to effect repression. In studies indicating that repression by IE2 requires direct bind- contrast, the ability of IE2 to repress transcription was dra- ing of IE2 to the crs element (11, 12). Quantitation of this matically decreased when the crs element was positioned 20 or experiment indicated that the concentration of IE2 required to 25 nt downstream from the TATA box (Fig. 2). IE2 had no repress 50% of the initial activity was approximately 10 nM, effect on transcription of a control template (Fig. 2) and, corresponding to a molar ratio of five molecules of IE2 per moreover, IE2 did not repress transcription of templates molecule of template. containing the crs element located 10 bp upstream of the IE2 efficiently repressed both TFIID- and TBP-directed TATA box (data not shown). The fact that repression was basal transcription. Quantitation of this experiment indicated dependent on the position of the crs element suggests that the that the extent of repression by IE2 was similar in both mechanism of repression might involve direct competition conditions (Fig. 1C). Since repression does not require the between IE2 and a component of the basal complex for binding TBP associated factor (TAF)-containing TFIID complex, yet to the HCMV MIEP. C A_-29 -23 -13 crsat l GGTCTATATAAGCAGAGCTcgttTAGTGAACCGTCACA Glcasete

-29 -23 -13 crsmut r GGTCTATATAAGCAGAGCTgcaaTAGTGAACCGTCACA Glss cassette| a TBP (Mutant crs) * TBP (Wild type crs) IID (Wild type crs)

0 20 40 60 80 B 1E2 (10-9 M) 1IE2 (pmol) - 0.15 0.75 1.5 3.75 - 0.15 0.75 1.5 3.75 - 0.15 0.75 1.5 3.75

*S*m

.4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

I I I I IID TBP

FIG. 1. IE2 represses transcription of the HCMV MIEP in transcription reactions reconstituted with either TFIID or TBP. (A) Diagram of the wild-type and mutant HCMV MIEP used in the in vitro transcription assays. The encompassing the crs element are underlined and the nucleotide substitutions are in lowercase type. (B) Transcription reactions were reconstituted. Purified recombinant IE2 protein was added to the reaction mixtures in the amounts indicated above the lanes. (C) Quantitative representation of the results shown in B. Downloaded by guest on September 29, 2021 2572 Biochemistry: Lee et al. Proc. Natl. Acad. Sci. USA 93 (1996)

IE IE2 IE2 IE IID,ItIA,IIB ID, IIA,I IB ID,lIIA IIA,IIB IID,I IA, IIB Factors added at O' complete Pll. IIF Pll Pl. IIF PlIilF 12 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1E2 (1.5 pmol at 30') + + _ + + + +

IE2 (1.5 pmol at0') +

1 3 4 5 6 7 * . 1 1

1 2 3 4 5 6 7 8 9 10 11 12 13

...... Mm: NTPs + Sark Stop

(0 30' 60' 90' I I I a I I I 1II 1I template + factors 4 pMIEP pMIEP pMIEP pMIEP missing factors (-1145/-7) (TATA 5 crs) (TATA 10 crs) (TATA 15 crs) +/- IE2 \TCGCGGGCCCGATC pMIEP (TATA 15 -7 crs) FIG. 3. Preinitiation complex intermediate containing stably bound RNA polymerase II is refractory to repression by IE2. Tran- scription reactions were assembled as outlined (Lower). (Upper) Bracketed lanes indicate reactions that received the same combination of factors at time 0. The reactions that received IE2 (1.5 pmol) at time O or after 30 min of incubation are indicated by the + symbol. Sark, sarkosyl. -29 -23 -13 -3

FIG. 2. Repression by IE2 is dependent on the position of the crs results suggest that initiation-competent complexes are refrac- element. (Lower) A schematic representation of the HCMV MIEP tory to repression by IE2 and that IE2 might block an plasmids containing oligonucleotide insertions between the TATA box intermediate step in complex assembly. and the crs element of the HCMV MIEP. (Upper) The HCMV MIEP Next, we analyzed the effect of IE2 on transcription after constructs were linearized with the restriction endonuclease Pvu II and different preinitiation complex intermediates were allowed to the products of the transcription reactions were analyzed with a run-off form. Addition of IE2 at 30 min had no effect on transcription assay. The solid arrowheads indicate the migration positions of the when only factors IIE and IIH were omitted from the initial RNA products transcribed from the wild-type and mutant test tem- incubation (Fig. 3, lanes 4 and 5), suggesting that preinitiation plates. The open arrowhead indicates the position of the RNA products transcribed from an internal control template. (Lower) The complex intermediates formed by TFIID, TFIIA, TFIIB, RNA DNA template utilized in each set of reactions is indicated. polymerase II, and TFIIF were resistant to inhibition by IE2. These conditions support the formation of the TFIID-TFIIA- Binding of IE2 to the crs Element Blocks the Recruitment TFIIB (DAB)-polymerase F (PolF) complex intermediate, of RNA Polymerase II to the Promoter. Kinetic, template which contains all four factors stably bound to the promoter commitment, and DNA binding experiments have shown that (ref. 23 and Fig. 4). Interestingly, reactions that only received transcription initiation in vitro is preceded by a preinitiation TFIID, TFIIA, TFIIB, and RNA polymerase II were also step in which the general factors and RNA polymerase II bind resistant to repression by IE2 (Fig. 3, lanes 6 and 7). In in an ordered fashion to the promoter (21, 22). By using crude contrast, preincubation of TFIID, TFIIA, and TFIIB was not nuclear extracts, we have shown (13) that repression by IE2 sufficient to prevent IE2-mediated repression (Fig. 3, lanes 12 resulted from competition of IE2 with the formation of a and 13). Furthermore, reactions that were preincubated with preinitiation complex intermediate. To confirm this possibil- a set of factors that would not support the formation of a ity, we carried out order-of-addition experiments designed to DAB-Pol preinitiation complex because either TFIIB or measure the effect of IE2 during the various stages of complex TFIID were omitted from the initial preincubation step were assembly. Transcription reactions were assembled in three sensitive to repression by IE2 (Fig. 3, lanes 8-11). Thus, these consecutive incubation steps as outlined in Fig. 3. In the first results strongly suggest that RNA polymerase II can associate step (at time 0), the DNA template and a complete set or a with the DAB complex and thereby prevent subsequent bind- subset of general factors and RNA polymerase II were mixed ing of IE2 to the crs element. Furthermore, these results imply and allowed to incubate for 30 min to allow the formation of that the crs element is accessible for IE2 binding in preinitia- preinitiation complexes. In the second step (at 30 min), the tion complex intermediates containing TFIID, TFIIA, and factors that were initially omitted were added with or without TFIIB. Identical results were obtained in experiments recon- IE2 and allowed to associate with the preinitiation complex for stituted with TBP instead of the TFIID fraction (data not an additional 30 min. In the third step (at 60 min), nucleotide shown). It is important to note that these experiments indicate triphosphates were added and the reactions were allowed to that the potential interactions between TFIID TAFs and elongate products for 30 min. The reactions were then stopped promoter sequences of the MIEP are not sufficient to coun- and analized. Sarkosyl was added 1 min after the nucleotide teract IE2 repression. The low levels of specific transcription triphosphates to a concentration of 0.2% to restrict the anal- observed in some reactions are probably due to inactivation of yses to one round of transcription. By using this protocol, we RNA polymerase II during preincubation conditions in which first analyzed transcription reactions containing a complete set the cannot associate stably with the promoter (Fig. 3, of basal factors at time 0 (Fig. 3, lanes 1-3) and found that, as compare lanes 4-7 with lanes 8-11). in Fig. 1, IE2 repressed transcription when added at time 0. In The Physical Association of IE2 and RNA Polymerase II contrast, IE2 had no effect on transcription when added after with the HCMV MIEP Is Mutually Exclusive. Our results 30 min of preincubation (Fig. 3, compare lanes 1 and 3). These suggest that repression by IE2 involves competition with RNA Downloaded by guest on September 29, 2021 Biochemistry: Lee et al. Proc. Natl. Acad. Sci. USA 93 (1996) 2573

0 C

LL N 0 LULU 0 CL U- N + LU cl i_ 0. CLL LU + + + + m mn m co c 0 EL a + + + f lql- m CL CLC LC cl IL CL N Co N m m EL E LU= mc LU Hn CL=- = L b, 4- D:B:PII:IIF 4- D:B:PII:IIF 1.

4 D:B:IE2p4O

IE2p4O -_ 4- D:B 4- D:B

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

FIG. 4. Binding of IE2 and RNA polymerase II to the HCMV MIEP is mutually exclusive. DNA binding reactions were performed with a DNA fragment containing the HCMV MIE core promoter with the wild-type (lanes 1-11) or a mutated (lanes 12-16) crs element. Factors were added as indicated above the lanes. The mobilities of the different DNA-protein complexes are indicated by solid arrows.

polymerase II for promoter binding. Alternatively, it is possi- polymerase 11 (23). Interestingly, addition of IE2p40 to binding ble that both RNA polymerase and IE2 bind the promoter but reactions containing TBP, TFIIB, TFIIF, and RNA polymer- without any consequence on transcription. To distinguish ase II resulted in the formation of DB-IE2p40 and DB-PolF between these two possibilities, we carried out the gel mobility complexes (lane 9). IE2p4O did not bind to a MIEP fragment shift experiments shown in Fig. 4. Binding of the full-length containing a mutated crs element (Fig. 4, lane 12) and could IE2 protein to the crs element produced a slowly migrating not supershift a DB complex formed on the mutated promoter complex (data not shown and ref. 11) that was difficult to fragment (Fig. 4, lanes 14 and 16). The concentrations of resolve from preinitiation complexes containing RNA poly- general factors and IE2p40 employed in the DNA binding merase II. Therefore, we decided to use a naturally occurring assay were 20- and 100-fold lower, respectively, than those used truncated version of IE2 (IE2p4O) that retains the DNA in the in vitro transcription reactions. Because the concentra- binding and autoregulation activity of the full-length protein. tion of IE2p40 used in the gel-shift experiments was not The role of IE2p40 during the of CMV infection sufficient to excert repression (data not shown) and the appears to be dedicated to downregulation of the HCMV concentration of factors were below saturation (i.e., the DNA MIEP. A histidine-tagged IE2p40 fusion protein was expressed probe was in excess), it is not surprising that IE2p40 did not in and purified from Escherichia coli. The activity of IE2p40 in prevent formation of the DB-PolF complex (Fig. 4, lane 9). DNA binding and repression was similar to that of the Unfortunately, the use of higher concentrations of IE2p40 in full-length IE2 protein (data not shown). Addition of IE2p40 the DNA binding reactions resulted in DNA-protein aggre- resulted in the formation of a specific DNA-protein complex gates that could not be resolved during electrophoresis (data (Fig. 4). In agreement with previous studies using the adeno- not shown). The fact that IE2p40 could supershift the DB virus major late promoter, binding reactions with TBP (lane 3), complex but not the DB-PolF complex indicates that IE2p40 TFIIB (data not shown), TFIIF (lane 11), or RNA polymerase has higher affinity for the DB complex than for the DB-PolF II (lane 10) did not produce specific DNA-protein complexes complex and free DNA and strongly suggest that the binding with a 32P-labeled fragment containing DNA sequences from of IE2 and RNA polymerase II to the promoter is mutually bp -34 to +5 of the HCMV MIEP. As expected, addition of exclusive. The most logical interpretation of these observations both TBP and IIB resulted in formation of the TFIID-TFIIB is that IE2 can cooccupy the MIEP with bound TBP and IIB (DB) complex (Fig. 4, lane 4) (24, 25). Addition of IE2p4O to and thereby prevents the association of RNA polymerase II binding reactions containing TBP and TFIIB almost com- with the DB complex. pletely shifted the DB complex to a slower migrating complex Heterologous DNA Binding Proteins Can Prevent the Re- (DB-IE2p40 complex). This observation indicates, consistent cruitment of RNA Polymerase II to the Promoter. To inves- with the functional data shown in Fig. 3, that IE2 can access tigate whether this mechanism of repression required specific a promoter associated with TBP and TFIIB. In agreement with properties in the repressor protein, we asked whether func- the previous experiments using the adenovirus major late tionally unrelated nuclear antigen EBNA-1 and BZLF proteins promoter, addition of both RNA polymerase II and TFIIF to of Epstein-Barr virus could function as transcriptional repres- binding reactions containing TBP and TFIIB resulted in the sors in the context of the HCMV MIEP. EBNA-1 plays a appearance of both the DB and the DB-PolF complexes (Fig. central role in the maintenance of Epstein-Barr virus latency 4). It has been shown that the DB-PolF complex contains both through its function in episomal DNA replication and trans- subunits of TFIIF and at least the largest subunit of RNA activation of latency (26, 27). BZLF (also known as zta, Downloaded by guest on September 29, 2021 2574 Biochemistry: Lee et aL Proc. Natl. Acad. Sci. USA 93 (1996) zebra, and EB1) is a member of the b-Zip family of transcrip- factors but not when added after functional preinitiation tion factors and regulates the expression ofviral genes required complexes were allowed to form (Fig. SB). As was the case with for the lytic replication cycle (28). IE2, EBNA-1 could not repress reactions that were preincu- Derivatives of the MIEP were prepared containing EBNA-1 bated with TFIID, TFIIA, TFIIB, and RNA polymerase II but and BZLF binding sites instead of the crs element. Addition of efficiently repressed reactions that were preincubated with a increasing amounts of BZLF-1 or EBNA-1 proteins had no subset of these factors (Fig. 5B). Identical results were ob- effect on transcription of the wild-type HCMV MIEP (Fig. tained in experiments reconstituted with TBP instead of the SA). When tested on promoters containing their cognate TFIID fraction (data not shown). These observations indicate binding sites in place of the crs, both BZLF and EBNAS9Ic that association of RNA polymerase II with the preinitiation proteins behaved as potent transcriptional repressors (Fig. complex blocks binding of EBNA-1 and that TFIID and TFIIB SA). The concentration of BZLF protein required to repress are sufficient to recruit the polymerase to the promoter. 50% of the initial transcription activity (IC50) was approxi- mately 10 nM (similar to the IE2 IC50). In contrast, the IC50 of EBNA,gic was 0.5 nM; therefore, 20 times more BZLF and DISCUSSION IE2 proteins were required to exert similar levels of repression. In this report, we have characterized the mechanism by which It is possible that this difference reflects the fact that the DNA the HCMV IE2 protein functions as a transcriptional repres- binding affinity for their cognate DNA binding sites is 20 times sor. Two lines of evidence suggested that repression results higher for EBNAI9Ic than for BZLF (10- I M and 2 x 10-9 M, from interference of IE2 with the binding of RNA polymerase respectively). Order-of-addition experiments as outlined in II to the preinitiation complex: (i) in vitro transcription exper- Fig. 3 indicated that EBNAl9lc (Fig. SB) and BZLF (data not iments showed that IE2 had no effect on transcription once shown) also repressed transcription by blocking RNA poly- RNA polymerase II was allowed to bind to the DB complex merase II binding to the promoter. EBNAl9lc repressed tran- (Fig. 3), and (ii) DNA binding experiments indicated that the scription of the HCMV MIEPEBNA-1 template when added physical association of IE2 and RNA polymerase II to the simultaneously with RNA polymerase II and the general promoter was mutually exclusive (Fig. 4). Additional evidence supporting this model comes from the experiment showing A that the position of the crs element is important for repression H-CMV MIEPBZLF H-CMV MIEPwt (Fig. 2). The of a 10-bp oligonucleotide between the

BZLF (pmol) - 0.5 1.0 2.0 4.0 0.5 1.0 2.0 4.0 TATA box and the crs element prevented repression by IE2 (Fig. 2). It has been well documented that in TATA containing promoters, the TATA box determines the transcription start 4w e0 as site of RNA polymerase II transcription in vitro (29). There- fore, the 10-bp insertion shifts the position of the crs element relative to the TATA box and the transcription start site but does not change the position at which RNA polymerase II 1 2 6 7 10 3 4 5 8 binds with respect to the TATA box. DNase I footprinting H-CMV MIEPEBNA1 H-CMV MIEPwt analyses with purified proteins have shown that simultaneous EBNA-1 (pmol) - 0.01 0.05 0.1 0.25 binding of TBP and IE2 results in the complete protection of promoter sequences upstream of the start site of the HCMV MIEP (11). When analyzed independently, the protection pattern of IE2 extended from bp -17 to +10 (10, 11) and that 11 1* *O-- of TBP extended from bp -35 to -15 with respect to start site of the HCMV MIEP (11). We propose that critical DNA contacts made by RNA polymerase II in promoter binding lie 11 12 13 14 15 within the region from bp -13 to -3 of the HCMV MIEP and that IE2 does not occupy this region when bound to a crs B IID.IIA IIDIIFA IlD,JIA element that is placed 10 bp further downstream. The fact that Factors added at O' complete IIB.PII IIB P1l IE2 cannot repress transcription from a crs element located EBNA- 1 (at 30') + between bp -3 and + 13 relative to the start site suggests that + + + once RNA polymerase II binds the promoter (to form the EBNA-1 (at 0 ) + DB-Pol or DB-PolF intermediates), the preinitiation complex becomes refractory to steric inhibition. If this observation is extrapolated to other promoters, it would suggest that repres- sors that bind downstream from the region between bp -13 and -3 of a TATA-containing promoter repress transcription of that promoter by a mechanism other than steric interference with the preinitiation complex. It is important to note that binding of IE2 to the promoter region between bp -3 and + 13 1 2 3 4 5 6 7 8 9 might disrupt TAF-DNA interactions. Our data does not rule out the possibility that disruption of TAF-DNA contacts might FIG. 5. Heterologous DNA binding proteins can block RNA play a critical role in this type of repression, if for example polymerase II recruitment. (A) In vitro transcription reactions showing these interactions are important for the stable association of the effect of BZLF and EBNA1l1c proteins in transcription from the the polymerase with the preinitiation complex in vivo. To our HCMV MIEP hybrid constructs. The DNA template and the amount knowledge the importance of TAF-DNA contacts on RNA of BZLF and EBNA191c proteins utilized in each set of reactions are polymerase II recruitment is not known. indicated. (B) Effect of EBNAi91c (0.1 pmol) in transcription reactions containing different preinitiation complex intermediates. Reactions Previously, eukaryotic transcriptional repressors that com- were assembled as outlined in Fig. 3 (Lower). The general factors pete with the general transcription machinery for promoter added at time 0 are indicated above each lane. The brackets group binding have been shown to block TFIID binding (4-6). reactions that contained the same set of factors at time 0. The + Several lines of evidence suggest that a repression strategy that symbol represents reactions that received EBNA1s1c (0.1 pmol) at depends on blocking TFIID binding would be most efficiently either 0 or 30 min as indicated. utilized for keeping transcriptionally inactive genes off rather Downloaded by guest on September 29, 2021 Biochemistry: Lee et al. Proc. Natl. Acad. Sci. USA 93 (1996) 2575 than for turning active genes off. These include in vitro 5. Kato, H., Horikoshi, M. & Roeder, R. G. (1991) Science 251, transcription experiments indicating that the TFIID complex 1476-1479. remains bound to the promoter after transcription initiation 6. Dostatni, N., Lambert, P. F., Sousa, R., Ham, J., Howley, P. M. (30) and that binding of TFIID to the TATA box is sufficient & Yaniv, M. (1991) Genes Dev. 5, 1657-1671. by (31), suggesting that 7. Pizzomo, M. C. & Hayward, G. S. (1990) J. Virol. 64,6154-6165. to prevent repression 8. Cherrington, J. M., Khoury, E. L. & Mocarski, E. S. (1991) J. transcriptionally active genes contain TFIID complexes stably Virol. 65, 887-896. bound to their promoters. Blockage of a step subsequent to 9. Liu, B., Hermiston, T. W. & Stinski, M. F. (1991) J. Virol. 65, TFIID binding may be the preferred strategy for interference 897-903. with actively transcribed genes, as the repressor would have 10. Lang, D. & Stamminger, T. (1993) J. Virol. 67, 323-331. access to its binding site each time the polymerase escaped 11. Jupp, R., Hoffmann, S., Depto, A., Stemberg, R. M., Ghazal, P. the promoter. This mechanism is suitable for the regulation & Nelson, J. A. (1993) J. Virol. 67, 5595-5604. of the HCMV MIEP as this promoter needs to be on early 12. Macias, M. P. & Stinski, M. F. (1993) Proc. Natl. Acad. Sci. USA in the infection cycle and off later in the cycle. 90, 707-711. It is surprising that among repressors that block the access 13. Wu, J., Jupp, R., Stemberg, R. M., Nelson, J. A. & Ghazal, P. (1993) J. Virol. 67, 7547-7555. of the preinitiation complex to DNA, IE2 represents the first 14. Lu, H., Flores, O., Weinmann, R. & Reinberg, D. (1991) Proc. example, to our knowledge, in which the repressor interferes Natl. Acad. Sci. USA 88, 10004-10008. with the RNA polymerase II recruitment step. Even though we 15. Reinberg, D. & Roeder, R. G. (1987) J. Biol. Chem. 262, 3310. did not analyze in detail the effect of core promoter sequences, 16. Flores, H., Ha, I. & Reinberg, D. (1990) J. Biol. Chem. 265, our data suggest that the only strict requirement for this type 5629-5634. of repression is the position of the repressor binding site. This 17. Flores, O., Lu, H. & Reinberg, D. (1992) J. Biol. Chem. 267, notion is further supported by the fact that two unrelated 2786-2793. factors that normally function as activators of transcription and 18. Peterson, M. G., Tanese, N., Pugh, B. F. & Tjian, R. (1990) have low and high DNA binding affinities function as potent Science 248, 1625-1630. repressors when their DNA binding sites are placed in a 19. Ha, I., Lane, W. & Reinberg, D. (1991) (London) 352, 689-695. configuration analogous to that of the crs element in the 20. Peterson, M. G., Inostroza, J., Maxon, M. E., Flores, O., Edmon, HCMV MIEP. Thus these data suggest that functional analyses A., Reinberg, D. & Tjian, R. (1991) Nature (London) 354, of core promoters and identification of relevant DNA binding 369-373. proteins will reveal new members of this class of eukaryotic 21. Zawel, L. & Reinberg, D. (1993) Prog. Nucleic Acids Res. Mol. repressors. Biol. 44, 67-108. 22. Conaway, R. C. & Conaway, J. W. (1993)Annu. Rev. Biochem. 62, We thank Kelly LaMarco, Danny Reinberg, Steve Mcknight, Greg 161-190. Peterson, Tim Hoey, and Nancy Stone for critical reading of the 23. Flores, O., Lu, H., Killen, M., Greenblatt, J., Burton, Z. F. & manuscript. We are grateful to Vijay Baichwal and Adam Park for Reinberg, D. (1991) Proc. Natl. Acad. Sci. USA 88, 9999-10004. providing EBNA-1 and BZLF proteins and to Gary Hayward for 24. Maldonado, E., Ha, I., Cortes, P., Weis, L. & Reinberg, D. (1990) providing the IE2cDNA. This work was supported in part by National Mol. . Biol. 10, 6335-6347. Institutes of Health Grants AI-30627 and MH-47680 (P.G.). P.G. is a 25. Buratowski, S., Hahn, S. & Sharp, P. A. (1989) Cell 56, 549-561. scholar of the American Leukemia Society. This is publication 8905- 26. Yates, J. L., Warren, N. & Sugden, B. (1985) Nature (London) IMM1 from the Scripps Research Institute. 313, 812-815. 27. Sugden, B. & Warren, N. (1989) J. Virol. 63, 2644-2649. 1. McKnight, S. L. & Yamamoto, K. R., eds. (1992) Transcriptional 28. Countryman, J. & Miller, G. (1985) Proc. Natl. Acad. Sci. USA Regulation (Cold Spring Harbor Lab. Press, Plainview, NY). 82, 4085-4089. 2. Tjian, R. (1981) Cell 26, 1-2. 29. Carcamo, J., Buckbinder, L. & Reinberg, D. (1991) Proc. Natl. 3. Hansen, U., Tenen, D. G., Livingston, D. M. & Sharp, P. A. Acad. Sci. USA 88, 8052-8056. (1981) Cell 27, 603-612. 30. Zawel, L., Kumar, P. & Reinberg, D. (1995) Genes Dev. 9, 4. Kaufman, P. D. & Rio, D. C. (1991) Proc. Natl. Acad. Sci. USA 1479-1490. 88, 2613-2617. 31. Workman, J. L. & Roeder, R. G. (1987) Cell 55, 613-622. Downloaded by guest on September 29, 2021